岚叔

统计数据

10

文章

0

粉丝

0

获赞

17

阅读

1周前

卧槽，朋友们，我已经分不清这是AI生成图片了。 Qwen-image-2512 这个模型可以的👍 提示词：一位中国女性大学生，性别女，年龄约20岁左右，超短发发型略带柔和文艺感，发丝自然垂落遮住部分脸颊，整体风格偏向假小子（tomboy）气质。她肤色冷白，五官清秀，表情略显羞涩又带着一丝拽劲，嘴角微微歪起，流露出痞帅又青春的神态。身穿一字领露肩短袖上衣，露出一侧肩膀，身材匀称。画面为近景自拍构图，人物占据主体位置，背景清晰可见宿舍环境：上铺床铺铺着白色床上三件套，旁边是一张桌面物品摆放整齐的书桌，以及木制柜子与抽屉。整体采用手机拍摄，光线为均匀柔和的环境光，色调自然真实，画面清晰明亮，氛围轻快鲜活，充满日常生活的青春气息。

2周前

儿时有个梦想就是山顶看场烟花秀借鉴了宝玉老师思路，提示词如下： { "template_name": "New Year City Skyline Celebration", "description": "A cinematic shot of a child overlooking a city skyline with custom fireworks effects.", "prompt_template": "A stunning cinematic shot of a joyful child in silhouette, wearing a whimsical party hat, standing on a high mountain peak with arms outstretched in wonder. The background is a breathtaking aerial night view of the Shanghai skyline, featuring the Oriental Pearl Tower, The Bund, and the winding Huangpu River under a starry night sky. The sky erupts with {city_fireworks_effect}. A massive, glowing firework display clearly spells out '2026' in the center. 8k resolution, photorealistic, magical atmosphere, highly detailed urban nightscape.", "input_variables": { "city_fireworks_effect": { "type": "string", "description": "Description of the style, shape, and vibe of the fireworks.", "default_value": "a spectacular Disney-style fireworks show, featuring Mickey Mouse shaped bursts, fairy-tale pastel colors, glittering magic dust, and dreamlike cascading sparks" } }, "full_prompt_example": "A stunning cinematic shot of a joyful child in silhouette, wearing a whimsical party hat, standing on a high mountain peak with arms outstretched in wonder. The background is a breathtaking aerial night view of the Shanghai skyline, featuring the Oriental Pearl Tower, The Bund, and the winding Huangpu River under a starry night sky. The sky erupts with a spectacular Disney-style fireworks show, featuring Mickey Mouse shaped bursts, fairy-tale pastel colors, glittering magic dust, and dreamlike cascading sparks. A massive, glowing firework display clearly spells out '2026' in the center. 8k resolution, photorealistic, magical atmosphere, highly detailed urban nightscape." } ``` 城市：夏威夷

2周前

opal 第一弹来了！为什么我认为opal潜力很大，我将在未来一段时间内展示10个案例来证明它的价值，并且将全部分享出来。 3分钟实现当天Hacker News 内容洞察：查询过去 24 小时内在 Hacker News 上发布的得分最高的前 10 篇文章。对于每篇文章，查询应指定检索其标题、URL、得分、ID 以及评论的完整文本。随后会使用Gemini Deep Research去获取实时数据最后整合内容提取高质量评论、形成网页，效果见截图，可以看到今天GLM-4.7模型信息也在榜上一个非常简单的页面应用就做好了，Gemini体验地址见评论👇

3周前

实操了下宝玉老师这个提示词，相当惊艳。绝对的年度MVP。针对宝玉第三步，我进行了优化，就是看看能否自动化。原第三步操作流：需要依次粘贴每一页 Slide 的内容描述。那有没有更快的方式？我进行了如下两种方式的尝试：方式一：使用谷歌新版Gem，也叫opal，这个可生成工作流或小应用。可以说完美适配宝玉这个多步能力，昨晚我就跑通了，但是发现一个问题就是目前opal 无法使用Gemini 3，pro版是不行。ultra账号我弄好了再试试，所以目前只能用上一代的生成出来，至少证明是跑通了。工作流分享链接放到评论里了方式一：目前普适度更高：第一步：就是在宝玉的gem跑完内容后，再追加一句话：所有内容放入json里 -- 这个目的是为了在生图窗口不用再逐个粘贴了第二步：在新窗口粘贴该json，最后附加上这句话：如上内容本对话的全局内容，需要全部记住。请根据内容先创建第一页ppt 每创建出一张，之后只需输入创建第二页、第n页即可效果见图二，用json还有个好处就是方便编辑内容啦以上内容均手打，觉得有用的朋友请点赞关注一波呀～

3周前

圣诞节快到了，来个圣诞主题，今年来点不一样自拍吧～玩法：可以文生图，也可以图生图（上传图片即可）动作可以自己来微调哈：修改这个字段 "panel_details": [ { "position": "左上", "action": "眨眼比出和平手势，手持拐杖糖" }, { "position": "右上", "action": "向镜头递出一份小包装礼物" }, { "position": "左下", "action": "比出手指爱心，同时吃着姜饼饼干" }, { "position": "正下", "action": "咯咯笑着，害羞地用手遮嘴，表情俏皮" }, { "position": "右下", "action": "回头看向身后，叠加“Merry Christmas”文字" } ] 提示词如下： { "template_name": "Christmas_Gift_Box_Inception_Collage", "description": "A surreal and festive 3D composition where the main subject interacts with a large physical prop (gift box) that contains a photo collage of themselves.", "visual_concept": { "style": "3D Photo-in-Real-Life / Meta-Photography", "effect": "Optical illusion where the collage appears printed inside a physical object", "atmosphere": "Warm, cozy, festive, playful" }, "composition": { "foreground_interaction": { "subject_pose": "Leaning comfortably on the top edge of the large gift box", "expression": "Warm, engaging smile, looking directly at viewer", "placement": "Top center, resting chin or arms on the box rim" }, "container_prop": { "type": "Large, vertical rectangular Gift Box", "state": "Open, standing upright", "material": "Deep red velvet texture with realistic pile depth, soft light absorption, and subtle directional sheen; avoid flat or synthetic-looking surfaces", "lid_position": "Leaning against the side or opened to the left/right to reveal content", "content_area": "The inside of the box displays the 5-panel photo collage — rendered as high-resolution, physically printed photos with visible paper texture, slight curl at edges, and natural lighting integration (no screen glow or digital overlay)" } }, "collage_content": { "grid_layout": "5-panel grid (Top: 2 square; Bottom: 3 vertical rects)", "inner_subject_styling": { "hair": "Short, dark brown bob", "accessory": "Red knitted reindeer antler headband", "clothing": "Cream colored ribbed knit cardigan" }, "panel_details": [ { "position": "Top Left", "action": "Winking, peace sign, holding candy cane" }, { "position": "Top Right", "action": "Offering a small wrapped gift to camera" }, { "position": "Bottom Left", "action": "Making finger heart shape, eating gingerbread cookie" }, { "position": "Bottom Center", "action": "Giggling, shy hand over mouth, playful expression" }, { "position": "Bottom Right", "action": "Looking back over shoulder, 'Merry Christmas' text overlay" } ], "render_quality_notes": "All collage panels must appear as authentic photographic prints — no pixelation, JPEG artifacts, compression blur, or artificial sharpening. Maintain consistent lighting direction and color temperature matching the main scene." }, "environment": { "setting": "Cozy living room floor", "background_elements": [ "Blurry Christmas tree with bokeh lights (warm yellow/gold)", "Scattered wrapped gifts in background", "Wooden floor texture with realistic grain and subtle wear marks" ], "depth_of_field": "Shallow focus (f/1.8), sharp on foreground subject and box, blurry background — but ensure background bokeh is smooth and natural, not digitally smeared or low-res" }, "color_palette": { "primary": "Christmas Red (Box, Antlers)", "secondary": "Cream/Off-White (Cardigan)", "accents": ["Gold (Ribbons, Lights)", "Pine Green (Tree)"], "lighting_temperature": "Warm (3200K - 4000K)" }, "technical_specs": { "camera_angle": "Eye level or slightly low angle to emphasize the size of the box", "lighting_setup": "Soft studio lighting, rim light on hair, warm ambient fill — with realistic shadow falloff and subsurface scattering on skin", "aspect_ratio": "3:4 (Portrait)", "quality_enhancements": { "anti_artifacts": "No banding, moiré, chromatic aberration, or halos", "texture_fidelity": "High-resolution textures for all materials (velvet, knit fabric, paper, wood) with micro-detail and proper scale", "photorealism_guidelines": "Avoid uncanny valley effects; maintain natural skin tones, eye reflections, and cloth draping physics. The collage must feel physically present inside the box — not pasted, floating, or screen-like." } } }

1个月前

睡前来点好看的：提示词（需要上传一张照片）： A meticulously crafted, vertically composed artistic Christmas portrait. A real person stands bathed in soft, warm holiday glow, cradling a transparent glass jar close to their face as if holding a gentle star. Inside the jar resides a miniature version of their younger self—the same individual—standing alone within a self-contained, snow-covered microcosm. These two "selves" are not mirror images; instead, their gazes intersect, initiating a silent dialogue between reality and distilled identity. The real figure is illuminated by gentle amber Christmas lights, their radiance dissolving softly into the background like a diffused constellation. The glass surface dances with fragmented reflections, splintering the true visage into layered shards that gradually fade along the jar’s curvature—forming a visual bridge between the outer world and the inner realm. Within the jar, snow is sculpted into evocative, lunar-like contours. Beside the tiny figure stands a solitary bottlebrush tree. The miniature self wears intricately detailed festive attire: woven textures, tufted wool fibers, and the plush tactility of winter garments are rendered with astonishing clarity. A serene luminescence glows from deep within the snow, casting an upward light upon the small figure—as if the snow itself still holds the memory of light. Frost patterns crystallize on the jar’s surface—not random snowflakes, but symbolic constellations echoing the twinkling lights behind the figure. Warm-toned fairy lights wrap around the vessel in abstract spirals, transforming it into a quiet cosmic nucleus. The real person’s expression is contemplative, almost cinematic in its depth. Their breath condenses into a delicate mist along the glass rim, adding a fleeting tenderness to the scene. The entire composition evokes layered universes: the external world lush and warm, the internal world refined, suspended, and sacred. For visual cohesion, costumes and props should remain consistent between the inner and outer figures. Rendered in 8K ultra-photorealistic quality, with masterful artistic lighting, sculptural reflections, and poetic textural detail—this image embodies an intimate, symbolic, and quietly supernatural Christmas aesthetic.

1个月前

周末给我家两岁宝宝用上了，没想到她出奇的喜欢！一直盯着屏幕，追着让我给她演示。通过实操发现，孩子对事物都很好奇，追着你告诉他这个是什么，那个是什么😂 她还会跟你学，这个增添了我们互动频率，玩的很开心所以感觉16:9 （适合大屏幕），物品越多越好，满足宝宝求知欲，之前也有推友标签标记不对，那么索性就去掉了标签箭头，避免误导，效果见图片。元提示词见评论。 PS：（晚点会再分享一个gemini 生图非常好用的免费平台，请持续关注🎉）

1个月前

肝了几小时，调出来一套适合2～5岁宝宝的：认知大发现图片：3D黏土风。8h后公布元提示词（因为还需要微调以确保质量），欢迎先点赞收藏，评论区留下你想要的风格和要求评论先留一个特定场景的提示词给大家先玩

#儿童认知 #3D黏土风 #教育资源 #2-5岁 #免费

1个月前

评论太多了，回复太多可能触达x 会话阈值了，打算直接公开元提示词了。留时间搞新创作了，请大家持续关注我的创作哈。（使用方式元提示词+场景生成具体场景的提示词，喂给nano banana即可）元提示词如下：最后动物园改成任意场景即可 # 核心配置 (Configuration) [SCENE_THEME] = {在此处填入场景，例如：海底世界 / 消防局 / 超市 / 恐龙公园} [TARGET_AGE] = 5-9 岁 (识字敏感期) # 绘图提示词 (Prompt Structure) 一张儿童识字小报《[SCENE_THEME]》，竖版 A4，学习小报版式，**整体加入柔和 3D 立体质感与适度景深，让画面更具沉浸性与空间层次**。 --- ## 一、小报标题区（顶部） 1. **大标题**：顶部居中展示标题：《[SCENE_THEME] 识字小报》。 2. **字体风格**：大字、醒目、卡通手写体、彩色描边。 * **附加要求**：标题文字呈现轻微 3D 浮雕质感，边缘有柔和光影，让标题在页面顶部更突出。 3. **装饰元素**：在标题周围添加与 [SCENE_THEME] 强相关的贴纸风装饰，颜色鲜艳。 * **附加要求**：装饰元素带轻微体积感，视觉上更像“贴在纸上的立体贴纸”。 --- ## 二、小报主体（中间主画面）画面中心是一幅卡通插画风的完整「[SCENE_THEME]」场景，**整体构图采用接近 2.5D–3D 效果的绘本风格**。 ### **整体气氛** 明亮、温暖、积极、色彩高饱和，同时加入柔光、景深、体积感，让画面具有空间深度与沉浸感。 ### **构图要求** 物体边界清晰，彼此之间留有贴标签的空白，避免拥挤。 * **附加要求**：前景—中景—后景分层明显，使标注对象更清晰，场景更立体。 ### **场景分区** * **核心区 A (活动)**：表现 [SCENE_THEME] 里的核心活动（孩子们或角色正在做什么），动作自然，带轻微 3D 立体造型感。 * **核心区 B (工具)**：展示场景特有物品或工具，物体具备简单的 3D 形体光影（非写实，而是绘本级别的立体感）。 * **核心区 C (环境)**：体现环境特征（建筑、植被、路标等），背景采用更低对比度与轻微景深，突出前景主体。 ### **引导角色** 1 位可爱的卡通人物（[SCENE_THEME] 的向导/工作人员），带轻微立体与阴影，使其具有温和的空间存在感，正在做欢迎或指引手势。 --- ## 三、必画物体与识字清单 (Generated Content) 请务必在画面中清晰绘制以下物体，所有物体采用 **清晰轮廓 + 色块分明 + 轻度 3D 塑形**： 1. **核心角色/设施 (Big Objects)**: {在此填入 4-5 个核心大物体，如：消防车、云梯、消防员} 2. **常见物品/工具 (Small Items)**: {在此填入 5-8 个具体小物品，如：水枪、灭火器、头盔、手套} 3. **环境/装饰 (Environment)**: {在此填入 3-5 个环境词，如：警报铃、消防栓、大门} --- ## 四、识字标注规则 (Labeling System) 对上述物体贴上中文识字标签： * **格式**：三行制 ① 拼音（带声调） ② 简体汉字 ③ 英文 * **样式**：彩色贴纸风（带轻微立体投影，让标签有“纸片贴上去”的空间感）。 * **位置**：紧靠对应物体，以引线连接，不遮挡主体。 --- ## 五、画风参数 (Style Parameters) * **风格**：Children's picture book style + Literacy tabloid aesthetic（儿童绘本风 + 识字小报风） * **附加风格要求**：Soft 3D shading, gentle depth-of-field, subtle volumetric lighting（柔和 3D、适度景深、轻体积光） * **色彩**：High Saturation, Warm Tone, Bright（高饱和、暖色、明亮） * **质量**：8k resolution, high detail, clean vector-like lines * **整体技术特征**：flat + soft 3D hybrid illustration（平面插画结合轻 3D 塑形）动物园

#元提示词 #儿童识字小报 #3D卡通场景 #识字教育 #绘本风格

1个月前

哈哈，对不住了锤哥，我大概推出来元提示词了。。顺便添加了个英文，有需要推导的元提示词的可以评论留言😂

#锤哥 #元提示词 #推导 #英文 #幽默